Serina Chang, UC Berkeley Electrical Engineering and Computer Science Assistant Professor, presented Generative AI and Statistical Methods to Simulate Fine-Grained Mobility Patterns from Aggregated Data at the Institute of Transportation Studies Transportation Seminar on Friday, March 6, 2026.
Abstract: Human mobility patterns are central to many policy domains, including transportation and public health. These patterns vary substantially across subpopulations and over time, yet capturing this heterogeneity is challenging due to data and modeling limitations. In this talk, I’ll describe our work using AI and statistical methods to infer fine-grained mobility patterns from aggregated location data. First, I’ll discuss our efforts to infer dynamic mobility networks during the COVID-19 pandemic, encoding hourly movements from neighborhoods to points-of-interest (POIs) with 5.4 billion edges. We integrated these networks into a model of COVID-19 spread, enabling new analyses of public health policies and health disparities. Second, I’ll present recent work on incorporating demographic heterogeneity into mobility trajectory generation with generative AI models. Since individual-level trajectories with ground-truth demographics are rarely available, we introduce a method that learns demographic-conditioned trajectories by combining unlabeled trajectories with region-level demographic compositions and aggregated mobility features. Overall, our work bridges real-world data constraints with the need for fine-grained mobility modeling, enabling more precise and equitable policies.
Bio: Serina Chang is an Assistant Professor at UC Berkeley, jointly appointed in EECS and Computational Precision Health and part of the Berkeley AI Research Lab (BAIR). Her research falls at the intersection of AI and human behavior, including modeling human behaviors with AI, evaluating and improving human-AI interaction, and developing AI tools for societal decision-making, with a focus on public health. Her work is recognized by the ACM SIGKDD Dissertation Award, KDD Best Paper Award, Forbes 30 under 30, Google Research Scholar Award, EECS Rising Stars, and Rising Stars in Data Science, and has received coverage from over 650 news outlets, including The New York Times and The Washington Post.